The Skills That Make You Irreplaceable in the Age of AI

---

🧠👨‍💻🪶 This Article, Even Though It Uses Emojis and Em-Dashes, Was Written Entirely by a Human Being.

---

The advent of AI (and Generative AI in particular) brings hype, uncertainty, excitement, and a wealth of opportunities to be explored.

With that, also comes the fear that AI will somehow replace certain roles or functions. This fear is certainly evident in the domain of software engineering where agents and copilots are now rampantly permeating all aspects of the software building process.

The curious thing is that the long-term question — “Can AI replace software engineers, architects, and others” is almost irrelevant at this point. Whatever the response to it may be.

The real issue is — 

Do organizations believe that they can decrease their technical workforce due to AI’s ability in taking over the applicable roles? 

This is the pertinent short-term to mid-term question and we are already starting to see it play out across the industry.

The short answer to it is yes — in the short term, an increasing number of companies believe that or at least are exploring the notion. In the long term, they will likely begin to realize that things are not as simple. In the short-term, however, it does seem that many organizations are looking for ways to leverage AI in replacing at least some of the work done currently by humans.

Whether all of this will decrease the need for certain roles remains to be seen. Frankly, it doesn’t feel to be case if we consider the long term needs of these organizations and of the industry as a whole. Many roles will change and become augmented by the capabilities of AI but the industry overall will still need human employees — and probably in an even higher number than today. That’s a topic for another day.

Whatever the case may be — both for the short and long term — there are skills that you can leverage to secure your role in this exciting and emerging world of increasingly capable systems.

The skills I am referring to are not rocket science. They are, for the most part, the same skill-set that has historically distinguished those who have enjoyed growth in their careers and those who did not. This is especially true for the field of technology and software engineering specifically.

The additional nuance is that, in the current climate, these skills become evermore important as we start sharing an increasing amount of our toil with machines who seem so eager to please, don’t get tired, and are full of sometimes awesome and sometimes unbelievably faulty ideas.

Here are three of these skills that will set you up for success in whichever role you are in and regardless of whether you want to “AI-proof” your career or simply want to make your career trajectory smoother, faster, and aimed higher.

Skill #1: Facilitation 🤝

When I tell software engineers and architects about the importance of this skill — many give me a look of bewilderment. “Isn’t this what project managers are for?” they ask.

Sure, PMs, and other roles that deal with the intersection of business and project flow, certainly bear their fair share of facilitating and organizing things. In fact, their role revolves around that. But that doesn’t mean that other roles can’t benefit from this skill.

What “Facilitation” means in this context is the ability to bring people together — typically in a formal setting such as a meeting or workshop — in order to produce a result. This goes far beyond scheduling a meeting and sending out an invite. If this is where it stops, that is the misfortune of the organization because facilitating an activity to success is an art as much as it is a science.

It requires the careful management of the meeting’s flow, pre-meeting preparation, post-meeting activities, and the orchestration of smooth collaboration between stakeholders to reach a goal in a successful and timely manner.

That is a mouthful but it is what it is. Learning how to effectively conduct meetings and to bring people together in various ways to achieve a common goal is often an underrated yet very powerful and prized skill in any organization.

The reason why this skill is so important for technical contributors such as software engineers and architects is because:

1. It sets you apart from your peers.

2. It makes you recognized and trusted on a whole different level by Business.

3. It gives you tremendous insight and control over a project than you would have otherwise. This, in turn, allows you to contribute a lot more value and that value is often a deciding factor in positioning you for career growth.

   
   

Skill #2: “Selling Ideas” 💡

Your Vibe-Coding buddy — whether it’s one of the heavy hitters such as GitHub Copilot, Tabnine, Gemini Code Assist, or one of the new kids on the block such as Kiro — it certainly awesome.

It can do a lot. It can serve as your trusted partner coding your application, auto-magically covering your repo with unit-tests (some working, some not — but that’s ok), and building robust (though often hard to follow) documentation.

If you are a software architect, GenAI can help you build diagrams with nifty tools such as Eraser AI, Edraw AI, and even ChatGPT. If you are curious how the most popular LLM stack up in Software Architecture activities, our article: Can I Replace Software Architects goes deep into it.

Not to mention AI workflow automation tools like n8n, AI design tools like Adobe Firefly, and a plethora of other useful GenAI and conversational AI apps that can augment and 10x your day-to-day work as technologist or someone working alongside of technologists (Product Owners, Product Managers, Analysts, etc)

This is all great and it saves significant time. However, one of the main things you will do — especially as you rise in your career as software engineer or architect — is to convince others of your ideas. In fact, this is something that you most likely do on an almost daily basis regardless of where you are in your career trajectory. 

It sometimes may go unnoticed, but this is definitely something that is part of the software developer’s, and especially that of an architect’s, work.

It may start with small, seemingly insignificant, ideas such as using this or that design pattern or the naming of variables. However, in time, as you progress in your career, it begins to touch ideas with long lasting and impactful decisions such as which technology to use, how to create the architecture of a large distributed system, or which cloud vendor the organization will open its wallet to.

You may be tempted to ask your favourite LLM and the LLM will gladly provide you with its take on things. But the LLM is eager to please, it gets confused fast, and it bases its decision on a limited and often generic set of decisioning criteria.

LLMs are excellent at finding gaps in your ideas, suggesting improvements, and pointing out things that you would have never noticed otherwise.

By the way, this is true for LLMs, SLMs, and Niche-specific LMs.

However, the AI will not be “selling the idea” for you. It will not go in convincing others in why the idea is good and why it should be implemented in that but not any other way.

That is something that is and will remain on you. The skills involved with selling an idea revolve around negotiation, communication, and the ability to convince others to join your frame of mind in exploring and eventually agreeing that yours is the best way forward.

AI can help fortify your idea or tear it down, but it won’t be your champion of pushing the idea forward.

Skill #3: Decision & Trade-Off Analysis 🚀

With al the talk about AI Agents and Agentic AI, somehow we miss the point that all what AI Agents are at their core is a language model with access to tools and some freedom to act independently. That is — independently of a human being who would otherwise be involved in triggering whatever behaviour the agent is exercising.

That’s it. 

Now, although many assume, and understandably so, that AI Agents are all about “agency”. That certainly is the idea behind the current movement and buzz around Agentic AI. That’s why so many organizations are rushing in headfirst into the abyss of POCs, pilot projects, and multi-million dollar strategic engagements with large consulting firms. All in effort to find ways of making agentic AI work and provide value.

AI agents have much promise for sure. The use cases are real and companies that are able to tap into the full potential of these agents are bound to reap some real value and position themselves at the frontier of digital innovation.

All that said, it’s going to be a long way, if at all until AI agents truly replace humans throughout any sizeable portion of the decision making chain. AI will augment, improve, help, and level up our ability to make and execute decisions.

But, the decision making process and the wider array of human activity that nurtures an idea from nothing and into a full fledged product, process, service — or anything elsle — goes far beyond what AI agents are able to do- at least in the current state of what is envisioned.

The analysis of decisions and trade-offs sounds like something that AI and LLMs in particular are well suited for. They certainly are. In fact, one of the emerging differentiators for

successful software architects and engineers is the ability to use LLMs effectively to “stress-test” decisions and trade-offs of various options.

Be careful though, even an LLM that has access to your entire organization’s knowledge and information repository, will struggle in making the right decision — even if it can lead you to it. Even when fed the right data and context, many decisions are still too multifaceted for an LLM. 

Moreover, decisions need varying degrees of trade-off analysis. Some decisions are insignificant enough not to warrant an analysis at all. Other decisions — warrant an extensive investigation, which can certainly be less time consuming with the help of AI, but still requires a human(s) to drive, question, and make any deductions with real impact on the organization.

Conclusion 

AI is a great augmenter and partner in your role as software engineer, architect, or technologist. But it has its limitations and some skill sets are way too multi-dimensional and complex for it to replace. At least for now and in the foreseable future.

Learn and hone these skills and you will future-proof your career.

In this article, we talked about three such skills:

1. Facilitation

2. Selling Ideas

3. Decision & Trade Off Analysis

There is more. If you are interested in learning about all the rest of the skills and how to future proof your career as a senior+ software engineer, software architect, or really any kind of technology architect — check out Unlocking the Career of Software Architect. It’s a comprehensive resource that reveals industry secrets for growth, success, and thriving in these roles.

Need Enterprise-Grade AI-Ready architectural templates?

Join the waitlist for the Architect’s Toolkit and be first to get the full set of templates when it launches.

👉 cloudwaydigital.com/architectstoolkit